sec11.html

11 Omega Support

(First available in 4.0N.)

This release of the previewer supports all aspects of Omega, an extended version of T_EX which supports enhanced multilingual capabilities and 16-bit font encodings. You can learn about the powerful capabilities of Omega, an opus of Yannis Haralambous and John Plaice, in TUGboat vol. 15, no. 3, September 1994, ``Progress on the Omega Project''. The Omega software is available at URL ftp://nef.ens.fr/pub/tex/yannis/omega, and there is a Web page at http://www.ens.fr/omega.

Omega is a tool for the multilingual expert, and a means for TRUET_EX to improve the performance of 8-bit T_EX on Windows. The plain T_EX and L^AT_EX macros are not directly usable in the Omega formatter, so the features are not immediately useful in the way an author would write conventional T_EX documents.

11.1 Extensions for Omega

TRUET_EX support for Omega includes the following features:

The previewer interprets DVI files containing character codes larger than a single byte.

The previewer loads 16-bit font metrics (Omega XFM or OFM files in addition to T_EX TFM files) and virtual fonts (Omega OVF or XVF format in addition to T_EX VF format).

Virtual fonts can directly access and render all code positions in 16-bit Unicode fonts such as are available in Windows NT and Windows 95. This means you can use any national character or math symbol anywhere in a Unicode font.

All extensions are upward compatible: the previewer still accepts the older TFM and VF files, as well as DVI files using the traditional single-byte encoding. The previewer will look first for an OFM or XFM file, and then a TFM file, when looking for a font metric file; and likewise for a OVF or XVF file, and then a VF file, when looking for a virtual font file.

11.2 Accessing Unicode Fonts

The most immediately useful aspect of the TRUET_EX Omega extensions are the simple and complete access to the Unicode fonts of Windows NT and Windows 95. This means that you can avoid encoding mismatches when using non-T_EX fonts in T_EX documents. For example, the Unicode Times New Roman font in Windows contains all of the characters (or elements needed to compose accented characters)¹ used in the T_EX Roman and L^AT_EX T1 encodings, and the Omega-enhanced TRUET_EX previewer allows full translation between T_EX and Unicode encodings via the Omega-enhanced virtual fonts. Access to Unicode allows us to finally rationalize the T_EX fonts and typesetting into cross-platform standards, instead of employing ad-hoc 8-bit encodings that will not work when, for example, the Windows code page changes or when an output file is exported via Adobe Acrobat.

While both the 16-bit and 32-bit editions of the previewer support the Omega files, rendering with Unicode fonts requires the 32-bit previewer running on a Unicode-capable version of Windows (currently Windows NT and Windows 95).

11.3 Sparse Font Optimization and Unicode Virtual Packets

(First available in 4.0P.) This release adds optimizations in the Unicode edition to optimize handling of sparsely populated fonts. For example, Unicode multilingual fonts may have characters mostly in the range 0x0000-0x3ff, plus a few others around 0x2000, and around the private zone starting at 0xfb00. The TRUET_EX sparse-font optimizations store compacted tables for these fonts, instead of populating 64 K-entry arrays with mostly empty positions. This avoid large memory demands which otherwise would result.

This release of the Unicode edition also implements virtual character packets for codes above 256; previously virtual characters larger than 256 could only appear in virtual fonts as codes for actual TrueType fonts; now TRUET_EX supports fully general T_EX and Omega virtual fonts. TRUET_EX also applies sparse table techniques to map virtual font character numbers to virtual font packets, thus avoiding inefficient use of memory.

HEAD